RESUMO
ObjectiveTo compare the efficacy between the Mycob.T scanner system and manual microscopy for detecting acid-fast bacilli in sputum specimens. MethodsBetween January and November 2020， a total of 1 519 sputum samples from suspected primary tuberculosis patients from 5 designated tuberculosis hospitals in Shanghai were examined by Smear and BACTEC MGIT 960 liquid culture （liquid culture） methods. Each specimen was subiected to 2 direct smear slides. One slide was stained by Z-N method and examined with manual microscopic method. Another slide was stained and scanned by the Mycob.T system. The efficacy of manual microscopy and the Mycob.T scanner system for detecting acid-fast bacilli in sputum specimens was compared based on the result of liquid culture. Results of the repetitive scanning by the Mycob.T scanner system and the recheck of the manual microscopy were analyzed. ResultsThe average positive rate by the Mycob.T scanner system was 14.4% （219/1 519） while the average positive rate by manual microscopy was 16.3% （248/1 519）. No significant difference was observed （χ2=2.13， P=0.145）. Based on liquid culture confirmation results， the sensitivity of manual microscopy （60.36%） was higher than that of the Mycob.T scanner system （52.94%）， and the difference is statistically significant （χ2=4.38， P=0.036）. Both methods had high specificity （98.94%）. The concordance of the Mycob.T scanner system and manual microscopy was 95.46%， with the kappa value of 0.826. The results of repeatability test of the Mycob.T scanner system and the recheck results of the manual microscopy showed that the coincidence rate of scanning by the Mycob.T scanner system was 99.5% （436/438）， and the recheck coincidence rate by the manual microscopy was 98.6% （432/438）. ConclusionThe Mycob.T scanner system have high specificity for detecting acid-fast bacilli in sputum samples and good consistency with the results of manual microscopy. Compared with manual microscopic examination， the Mycob.T scanner system can greatly alleviate the work intensity.